Available online at www.annclinlabsci.org 118 Annals of Clinical & Laboratory Science, vol. 42, no. 2, 2012 Immunohistochemical Evaluation of Inhibin-alpha in Non-Small-Cell Lung Carcinomas - a Pitfall in Diagnosing Metastatic Pulmonary Carcinomas Kai Zhang, Jianhui Shi, and Fan Lin Geisinger Medical Laboratory. Geisinger Medical Center, Danville, Pennsylvania Abstract. Context: Inhibin-alpha is commonly used in differentiating non-small-cell lung carcinoma from certain metastatic carcinomas, but a study on a large series of primary lung cancer cases has not yet been published. Objective: To establish whether non-small-cell lung carcinoma can express inhibin-alpha; if so, caution should be exercised when using the molecule to evaluate metastatic lung cancer in both lung and extrapulmonary sites. Design: 187 cases of non-small-cell lung carcinoma and small-cell lung carcinoma were evaluated for expression of inhibin-alpha on both routine and tissue microarray sections by immu- nohistochemistry. These cases included 90 mixed or acinar adenocarcinomas, 2 acinar carcinomas with mucinous feature, 6 bronchioloalveolar carcinomas, 9 large-cell carcinomas, 2 adenosquamous carcinomas, 2 sarcomatoid carcinomas, 41 squamous cell carcinomas, 12 typical carcinoid tumors, 3 atypical carcinoid tumors, 20 small-cell lung carcinomas, and 19 cases of normal lung. The staining intensity was graded as weak, intermediate, or strong. The distribution was recorded as negative (no staining), 1+ (<25%), 2+ (26- 50%), 3+ (51-75%), or 4+ (>75%). Cytoplasmic coarse granular staining for inhibin-alpha was the indica- tor for a positive result. Results: A subset of primary mixed or acinar and mucin-producing acinar carcino- mas (10%) and large-cell carcinomas (22%) expressed inhibin-alpha. No expression of inhibin alpha was observed in the remaining cancers or normal lung samples. Conclusion: Since a significant percentage of non-small-cell lung carcinoma cases expressed inhibin-alpha, caution should be taken when using inhibin- alpha as the key antibody for exclusion of a lung primary in lung and other organs. Introduction cells in mucinous cystic neoplasm of the pancreas and ovary, endometrial carcinoma, placental site Distinction of primary lung adenocarcinoma and tumor/nodule, prostate carcinoma, and hemangio- large-cell undifferentiated carcinoma from extra- blastoma of the brain[1, 3-15]. However, based on pulmonary carcinoma with similar histology meta- the literature, the expression of inhibin-alpha in static to lung is commonly encountered in surgical pulmonary carcinomas is known as a rarity. pathology[1,2]. Additionally, a differential diagno- Recently, our group and Tigrani and colleagues, sis between a primary adrenocortical neoplasm independently observed by immunohistochemistry and an adrenal metastasis from the lung is fre- that a significant percentage of non-small-cell lung quently raised during lung cancer staging. Inhibin- carcinomas (NSCLC) of the lung could express alpha, a glycoprotein, is a highly sensitive and spe- inhibin-alpha [16, 17]. To further confirm the find- cific marker for identifying adrenal cortical ing, immunohistochemical detection of inhibin- neoplasms and gonadal sex-cord stromal tumors. alpha expression was performed for a large series of Expression of inhibin-alpha has also been reported NSCLC and neuroendocrine carcinomas of the in other benign and malignant neoplasms, such as lung. pancreatic serous microcystic adenoma, stromal Materials and Methods Address correspondence to Kai Zhang, Geisinger Medical Laboratory. Geisinger Medical Center, Danville, Pennsylvania; 100 N. Academy 187 cases of primary lung carcinomas or tumors Ave. Danville, PA 17822; tel: 570-271-6332; fax: 570-271-6105; email: [email protected] from 187 patients in a 5-year period (2005-2009) 0091-7370/12/0200-118. © 2012 by the Association of Clinical Scientists, Inc. Immunohistochemical Evaluation of Inhibin-alpha 119 Table 1. Immunohistochemical staining results for inhibin-alpha in 19 cases of normal lung and 187 cases of primary pulmonary carcinoma Diagnosis Positive Cases Total % of Positivity 0 1+ 2+ 3+ 4+ Normal lung (N=19) 19 0% Mixed or acinar AC (N=90) 81 4 4 1 0 10% (9/90) Non-mucinous BAC (N=2) 2 0% (0/2) Mucinous BAC (N=4) 4 0% (0/4) AC with mucinous feature (N=2) 1 1 50% (1/2) LCC (N=9) 7 1 1 22% (2/9) Adenosquamous carcinoma (N=2) 2 0% (0/2) SCC (N=41) 41 0% (0/41) Sarcomatoid carcinoma (N=2) 2 0% (0/2) Typical carcinoid (N=12) 12 0% (0/12) Atypical carcinoid (N=3) 3 0% (0/3) Small cell carcinoma (N=20) 20 A total of 187 lung cancer cases with 12 /187 (6.4%) cases positive for inhibin-alpha AC-adenocarcinoma. LCC-large cell carcinoma. BAC-bronchioloalveolar carcinoma. SCC-squamous cell carcino- ma. Immunoreactivity grading: 0 (no staining), 1+ (<25%), 2+ (26-50%), 3+ (51-75%), or 4+ (>75%). were retrieved from the archives of the authors’ in- neuroendocrine carcinoma (LCNEC), 2 cases of stitution. The study was approved by the sarcomatoid carcinoma, 35 cases of squamous cell Institutional Review Board at Geisinger Medical carcinoma, 2 cases of adenosquamous carcinoma, Center. All cases, including H&E stained slides and 40 cases of typical, atypical carcinoid tumor, and immunostained slides (cytokeratin [CK]-AE1/ and small-cell lung carcinoma. All cases of TMA AE3, CK7, CK20, CK5/6, p63, MOC-31, cal- were constructed with duplicate 1.5 mm tissue retinin, WT1, S-100, pCEA [polyclonal carcino- cores as previously described [18]. embryonic antigen], CDX2, CD56, synaptophy- sin, chromogranin, vimentin, desmin and Immunohistochemical stains were performed on myogenin) were re-evaluated to ensure correct di- sections of TMA sections and /or the routine paraf- agnosis and classified according to the criteria of fin embedded blocks of the tumor and normal lung the 2004 World Health Organization Classification previously described [18]. The tissue sections were of Tumours of the Lung and Pleura, Thymus and deparaffined. Antigen retrieval conditions included Heart [2]. heat-induced epitope retrieval in DAKO Histologic diagnosis of 187 cases (96 male, 70 fe- Cytomation Target Retrieval Solution (code no. male; age range, 34-88 years; mean, 62.4 years) S1700) for 30 minutes. The tissue sections were in- indicated 90 cases of mixed-subtype or acinar ade- cubated in 3% hydrogen peroxidase for 5 minutes nocarcinoma (16 cases of resected samples and 74 to quench endogenous tissue peroxidase. The tissue cases in tissue microarrays [TMA]); the remaining sections were then incubated with a mouse mono- cases were in TMAs, including 6 cases of bronchio- clonal antibody against human inhibin-alpha loalveolar carcinoma (BAC) (4 non-mucinous and (Serotec, MCA951ST, IgG2a, clone R1) at a 1:80 2 mucinous BAC), 2 cases of adenocarcinoma with dilution for 30 minutes at room temperature. The mucinous features, 9 cases of large cell carcinoma slides were stained in an automated immunostainer (LCC) including 3 cases of large cell using a DAKO EnVision HRP kit. 120 Annals of Clinical & Laboratory Science, vol. 42, no. 2, 2012 Figure A. Normal lung negative for inhibin-alpha (x 400). Figure B. Mixed adenocarcinoma showing 3+ immunoreactivity for inhibin-alpha (x 400). Figure C. Adenocarcinoma with mucinous feature showing 1+ immunoreactivity for inhibin-alpha (x400). Figure D. Large cell carcinoma showing 4+ immunoreactivity for inhibin-alpha (x 400). Immunohistochemical reactions were developed Results with diaminobenzidine (DAB) as the chromogenic peroxidase substrate, and the slides were counter- Immunohistochemical results are summarized in stained with hematoxylin. The adrenal tissue and a Table 1. No expression of inhibin-alpha was ob- case of adrenal carcinoma served as positive con- served in 19 cases of normal lung tissue (Figure A). trols. Negative controls included replacement of A total of 187 cases of primary lung carcinomas the primary antibody with nonimmune serum. were evaluated; 12 of 187 (6.4%) of the lung can- The cells showing visible cytoplasmic staining for cers were positive for inhibin-alpha in this study. inhibin-alpha were recorded as positive. The stain- The tumor cells showed cytoplasmic granular ing intensity was then subjectively graded as weak, staining with high intensity in all positive cases. intermediate, or strong. The distribution was re- The distribution of the staining showed varied corded as negative (no staining), 1+ (<25% of tu- grades, ranging from 1+ to 4+ positivity. The posi- mor cells stained), 2+ (26-50% of tumor cells tive cases were limited to adenocarcinoma and stained), 3+ (51-75% of tumor cells stained), or 4+ large cell carcinoma. Nine of 90 (10%) cases of (>75% of tumor cells stained). mixed or acinar adenocarcinomas, including 3 of Immunohistochemical Evaluation of Inhibin-alpha 121 16 cases in routine large tissue sections and 6 of 74 high tumor grade was not demonstrated convinc- cases in the TMA sections, were positive for inhib- ingly in our current study. in-alpha (Figure B). Among the 9 positive cases of The intensity (weak, intermediate, and strong) and mixed or acinar adenocarcinoma, 1 case of mixed distribution (1+ to 4+) of inhibin-alpha staining do acinar and solid showed 3+ positivity; 4 cases, in- not show any particular association with tumor cluding 2 cases of mixed acinar and solid pattern, 1 growth pattern, tumor size, or tumor grade. Coarse, mixed acinar and lipidic, and 1 mixed acinar and granular cytoplasmic staining is the only staining papillary showed 2+ positivity; and 4 cases, includ- pattern observed which is similar to adrenal corti- ing 3 cases of acinar and lipidic pattern and 1 cal neoplasms and gonadal sex-cord stromal mixed acinar and lipidic pattern, showed 1+ posi- tumors. tivity. There was no obvious association among It is well accepted that inhibin-alpha expresses in positive cases or between intensity of positive stain- the majority of adrenal cortical neoplasms (ACNs), ing and varied growth pattern of mixed adenocar- and ACNs frequently express Mart-1 and calretinin cinoma.